Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

The author(s) shown below used Federal funds provided by the U.S. Department of Justice and prepared the following final report:

Document Title: Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Author(s): Alissa Pollitz Worden

Document No.: 199912

Date Received: September 2003

Award Number: 98-WT-VX-K001

This report has not been published by the U.S. Department of Justice. To provide better customer service, NCJRS has made this Federally-funded grant final report available electronically in addition to traditional paper copies.

Opinions or points of view expressed are those of the author(s) and do not necessarily reflect the official position or policies of the U.S. Department of Justice. Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

By Alissa Pollitz Worden

December 2000 NCJ 199912 Alissa Pollitz Worden, Ph.D., is with the School of Criminal Justice, University at Albany, State University of New York.

Findings and conclusions of the research reported here are those of the authors and do not reflect the official position of the U.S. Department of Justice.

The project directors were Alissa Pollitz Worden, Ph.D., and Bonnie E. Carlson, Ph.D., CSW, both of whom are at the University at Albany, State University of New York. Dr. Worden is with the School of Criminal Justice; Dr. Carlson is with the School of Social Welfare. The research was supported by the National Institute of Justice (NIJ) under grant number 98–WT–VX–K011 with funds provided under the Violence Against Women Act.

The National Institute of Justice is the research, development, and evaluation agency of the U.S. Department of Justice and is solely dedicated to researching crime control and justice issues. NIJ provides objective, independent, evidence-based knowledge and tools to enhance the administration of justice and public safety.

The National Institute of Justice is a component of the Office of Justice Programs, which also includes the Bureau of Justice Assistance, the Bureau of Justice Statistics, the Office of Juvenile Justice and Delinquency Prevention, and the Office for Victims of Crime. Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

By the early 1990s, the limitations of traditional criminal justice responses to violence against women, particularly domestic violence, were recognized by most practitioners and policymakers. Criminal justice agents were faulted for not using their power to enforce the law frequently or consistently enough to hold offenders accountable. In addition, the laws appeared inadequate to protect victims from violence that often occurred without witnesses and without unambiguous evidence. The criminal justice process, fragmented and sequenced, appeared to be more of an obstacle course than a refuge for victimized women.

Beginning in the 1980s, these concerns prompted diverse remedies: better law enforcement training, tougher arrest policies, and greater victim advocacy in the courthouse. But more recently the focus of reform has shifted toward purposeful attempts to coordinate responses across agencies and across the domains of criminal justice, social services, and victim advocacy. Through the initiative and leadership of local task forces and coalitions, efforts to coordinate community responses are now widely endorsed as the most promising strategy for long-term change.1 However, the people who work together on these strategies, although sharing a common concern about improving responses to domestic violence, seldom share a common knowledge base. They may be familiar with conventional wisdom about the performance of criminal justice agents and the behavior of victims and offenders, but they are not familiar with research that examines these common assumptions; they often have little access to research that evaluates the programmatic and policy initiatives that are placed on their agendas; and they have little opportunity to learn about the experiences of other communities that have implemented coordinated intervention strategies and the impacts of those strategies. Importantly, they may know little about the perspectives, attitudes, and professional roles of those with whom they seek to collaborate. This report attempts to address these information needs, and is directed to the diverse people who are drawn together to work on local coordination efforts.

Participants in task forces and coalitions face many challenges: They must assess the nature and magnitude of violence against women in their communities; they must analyze local systems to understand who encounters criminal justice and social services agencies, under what circumstances, and with what results; they must draw inferences about how to improve local responses, or create new ones, from their understanding of others’ experiences; and they must assist in implementing and sustaining changes, sometimes in the face of resistance or skepticism from many quarters. Confronting a fragmented criminal justice system, and often a lack of communi- cation between criminal justice agents and other community organizations, they typically begin their work with the hope that organizing a few, or many, elements of the community’s response around a common objective will produce better results than merely adjusting specific practices or creating new programs.

This report summarizes social science research in three broad topic areas that are likely to be of interest to local task force members and leadership:

Conventional wisdom about domestic violence, offenders, and victims. Attempts to change and improve conditions usually begin with some commonsense assumptions about the targets and beneficiaries of the change. These assumptions may be grounded in experience,

1 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

but sometimes they are passed on simply because they are plausible and commonly accepted. It is important to scrutinize practitioner folklore about domestic violence, as well as common beliefs about victim and offender behavior.

Research on criminal justice reforms. Many innovations within, across, and external to criminal justice agencies have been the subject of evaluation research. This research suggests that policy changes do not always consistently produce the results that were predicted or hoped for, but also that some of the more recent and promising innovations have not yet been adopted and evaluated to the extent that would permit conclusions to be drawn about their effectiveness.

Research on the initiation, implementation, and impact of community-level change efforts. A small but illuminating body of research documents the diversity of efforts to establish and maintain coordinated community intervention strategies, and begins to describe the conditions under which they emerge, persevere, or dissolve. A smaller but important set of studies has examined the impact of coordinated community responses on important outcome measures, such as recidivism rates. This research suggests that investments in long-term strategies aimed at long-term goals, such as intra- and interagency changes in practice and policy, should be accompanied by realistic expectations about the pace of change and progress toward important long-term goals.

Revisiting and Revising Conventional Wisdom About Domestic Violence, Victims, and Offenders

The legal and social history of violence against women in the United States is probably familiar, at least in its general outlines, to most community agents involved in local domestic violence intervention policy. Physical violence inflicted on women by their husbands was de facto legally permissible in many States, inasmuch as many courts in the 19th century exempted husbands from assault statutes and legislatures exempted them from rape statutes (Bonsignore et al., 1989). These laws changed slowly and unevenly across the States (Ryan, 1996). By the 1960s and 1970s, grassroots organizations that provided shelter and support to women victims also began advocating for systemic and policy changes, challenging widely held beliefs that domestic violence resulted from stress, marital conflict, or women’s provocation. Instead, they maintained that a culture of male privilege and dominance served to rationalize and excuse violent behavior in personal relationships, and that culture was supported or at least tolerated by the legal system (Schechter, 1982; Pence, 1983; Pleck, 1987). Much of this advocacy was aimed at local police practices, as the mission of advocates expanded to include not only helping victims, but holding offenders accountable for criminal behavior as well (Davis, Kunreuther, and Connick, 1984).

By the 1990s, most laws that exempted or justified violence against women had been replaced by more gender-neutral definitions of crime (in the case of sexual assault), laws that expressly prohibited forms of violence that formerly were tolerated (such as marital rape), and even laws that created new offenses to target offenders whose victims were predominantly women (such as stalking statutes).2 Police practices and policies that expressly discouraged enforcement

2 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

responses (such as arrest) in favor of on-scene mediation and reconciliation had been largely discarded (Buzawa and Buzawa, 1996). However, increased interest in and concern about violence against women also led to important research findings:

Partner violence is more common than often believed.3

Criminal justice agencies encounter only about 25 percent of partner assault cases.

Many offenders commit multiple forms of abuse, and recidivate frequently (Tjaden and Thoennes, 2000).

It appeared that even with substantial changes in statutory protections, many victims could not find safety, support, or resources to stop the violence in their lives.

This unsettled state of law and basic knowledge about violence may explain why many practitioners subscribe to unfounded beliefs about the behavior of victims and offenders, beliefs that may frustrate collaborative efforts to change local policy and practices (see Worden et al., 1999). These beliefs include negative stereotyping of victims as “uncooperative” and pessimism about the effectiveness of strategies aimed at changing offender behavior—attitudes that, when combined, place undue reliance on victims to protect themselves from violence over which they may have little control.

The criminal justice process is primarily reactive: It responds to cases that are brought to its attention, and it usually discards those whose complainants fail to perform the roles expected of them by prosecutors and judges. In cases of domestic violence, court actors have historically relied on the victim to “make the case,” so a reluctant, frightened, or ambivalent victim is often labeled “uncooperative.” Interviews with knowledgeable criminal justice practitioners reveal a common theme: The chief source of frustration and concern in handling domestic cases is the “reluctant,” “recanting,” or “uncooperative” victim, who not only fails to follow through with the criminal process, but also reunites with her abuser (Worden et al., 1999).

Among those victims who do seek help from the legal system, an unknown, but probably not trivial, percentage request that charges be dropped; another unknown number prefer not to participate actively in legal proceedings, although they may not wish to have those proceedings terminated. Most prosecutors do not record these preferences in their files or in court records, so researchers cannot know how many victims feel this way. Moreover, the two issues—withdrawing from prosecution and returning to abusive partners—are frequently spoken of simultaneously by practitioners (who may assume that ending the court case is a step in the direction of reconciliation), but research reveals that they should be analyzed separately. Shelter samples indicate that, contrary to conventional wisdom, more than half of the women who use shelters ultimately terminate the relationship (Sullivan et al., 1994). Moreover, recent research reveals that many women in abusive relationships are realistic and pragmatic about the economic and social barriers, as well as the potential for escalated violence, that they face if they attempt to leave

3 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

(Sullivan et al., 1992a, 1992b). Women who are economically independent of partners are more likely to terminate abusive relationships (Sullivan et al., 1994).

Furthermore, research on the motivations of women involved in domestic violence prosecutions reveals that victims have diverse and rational explanations for both pursuing and, sometimes, terminating prosecution—reasons that included securing personal safety, ensuring protection and support for children, getting abusive partners into treatment or therapy, and simply ending the case because it appeared that the violence had ended (Ford, 1991). It is important to note that these explanations for withdrawing do not imply irrational reconciliation with batterers, nor do they differ much from the reasons other misdemeanor complainants give for deciding not to pursue conviction (Merry, 1990).

As practitioners accept greater responsibility for intervening in violence against women, and as they better understand the impact of violence on victims, they increasingly recognize that lenient or casual responses to perpetrators are inappropriate. However, responses that might deter, rehabilitate, or incapacitate offenders are costly and punitive, much more so than those that might be imposed on other misdemeanants. Besides, it is not at all clear what, if anything, affects future criminal behavior. It is far easier to turn attention to the victim, by investing resources in safety planning, social services, and protection; a strategy known as “target hardening” in other criminal justice venues.

This is an understandable adaptation to a seemingly impossible challenge. Studies reveal that recidivism rates in domestic violence incidents are high, regardless of how they are measured. The most conservative measure, rearrest, yields a rate of 20 percent in some studies (Steinman, 1988; Murphy, Musser, and Maton, 1998). Victim interview data suggest significantly higher rates of 40 percent or more (Shepard, 1992; Syers and Edleson, 1992), and in some studies the rate of a subsequent assault is as high as 80 percent (Garner, Fagan, and Maxwell, 1995). It is safe to say that among women who seek help from the police, there is a better than even chance of being assaulted again within 6 months. If the goal of criminal justice is to reduce the chances of future violence, the most promising strategy at this point may be protecting the victim. However, research on men who are abusive and assaultive in relationships reveals that some offenders desist over time, even without formal interventions (Fagan, 1989); that some characteristics of offenders are associated with desistance following legal intervention such as arrest (Sherman et al., 1992; Hirschel and Hutchison, 1992; Garner, Fagan, and Maxwell, 1995); that, more generally, men’s victimization of women is associated not only with moral values but also with rational calculations about consequences (Williams and Hawkins, 1992); and that no single model accurately distinguishes men who are abusive and violent from those who are not (Saunders, 1992). (See Saunders and Hamill, 2003, for a discussion of research on offender interventions and treatment.)

In summary, conventional wisdom about violence, victims, and offenders has frequently served to justify longstanding policies or practices, and like most conventional wisdom, these assumptions may carry some measure of truth. However, research suggests that such assumptions should be scrutinized carefully, inasmuch as they usually oversimplify complex issues, often

4 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

overgeneralize from limited experiences, and, when incorporated into practice, can become self- fulfilling prophecies. Where victims are treated with skepticism, they are likely to distrust and withdraw from systems intended to protect them, and where perpetrators are seen as incorrigible, they are unlikely to be routed into programs aimed at changing their behavior.

The Effectiveness of Criminal Justice Reforms

Communities have witnessed a proliferation of reforms and new policies in the area of domestic violence, many of them directed toward criminal justice agency performance. Many of these innovations have been adopted at the initiative of local task forces, with the objective of reducing violence in the community or among a subset of offenders, increasing victims’ safety or sense of security, or increasing the community’s capacity to hold offenders accountable for their actions. Most reforms involve fairly small, incremental changes in practice, although the goals that they seek are often broadly defined and optimistic.

Research on these interventions and reforms is of two general types. Process evaluations, which are frequently required by external funding sources for new programs, objectively document the implementation of a new program or policy. They are intended to inform readers about how the program was adopted by and adapted to a local environment. Outcome evaluations are systematic tests of the predictions, or hypotheses, that underlie the policy change or program. Outcome evaluations provide empirical assessments of the policy or program’s effect on desired (and sometimes unexpected) outcome variables, such as recidivism, compliance with treatment, or conviction rate. The answers to the frequently asked question, “What works?” will be found, if it is to be found at all, in these kinds of studies. The following sections briefly summarize the findings of research of particular interest to practitioners involved in community-level innovations.

Police Practices and Policies For many, arrest is synonymous with law enforcement, and police were widely criticized in the 1980s and 1990s for failing to arrest offenders in domestic violence cases (Ferraro, 1989). While it is true that police training and policies in the 1960s and 1970s explicitly discouraged arrest and encouraged officers to mediate or simply defuse what were seen as domestic conflicts (Parnas, 1967; Bard and Zacker, 1971), few if any police departments would publicly espouse such a policy today. In most States, police have authority to make arrests in domestic violence situations when they have probable cause to believe a misdemeanor-level offense has taken place, which is an exception to standard misdemeanor arrest law.4 In a few States, police are legally required to make arrests under specified circumstances. However, legal and empirical research reveals that police still retain considerable discretion in making arrests, because officers still must make judgments about the legal character of the offense. As a result, arrest rates vary greatly across communities, but arrest remains an unlikely outcome in most incidents. For example, a recent study of five New York communities revealed that arrest rates for recorded incidents ranged from 25 to 40 percent (Worden, 2000a). It is also important to note that although arrest rates are typically reported as the percentage of cases in which police file written reports, report writing is neither mandatory nor universal in many departments.

5 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

The question of whether police are making as many arrests as they should is both a normative and an empirical one. If the standard for domestic cases is the arrest rate in similar offenses involving nonfamily acquaintances or strangers, then some evidence suggests that rates are comparable, and comparably low, since arrests are uncommon in assaults involving bar fights and neighbor disputes as well (Dutton, 1988). Some research suggests, however, that police are less likely to arrest in domestic incidents, all other factors being equal (Fyfe, Klinger, and Flavin, 1997). Some would argue that this standard is inappropriate inasmuch as domestic violence victims are at greater risk than most other assault victims of repeat victimization when no arrest is made.5

Further, although 40 to 50 percent of all cases to which police are called involve off-scene suspects (Erez, 1986; Worden, 2000a), early research suggested that in domestic incidents police seldom pursue suspects who flee the scene before police arrive. Off-scene pursuits were so uncommon in the diverse cities studied in the well-known spouse assault arrest experiments in the late 1980s that those cases were excluded from the research altogether (Sherman et al., 1992). However, a recent study of upstate New York police departments suggested that rates for off- scene pursuit were almost equal to those for on-scene arrests in communities with strong department-level support for arrest policies (Worden, 2000a).

More extensive use of arrest in domestic violence cases has been promoted for three reasons. First, arrest is thought to constitute an unambiguous message of condemnation to offenders—and to society more generally—a corrective to earlier practices that condoned family violence. Some argue that arrest serves a symbolic function, rightly pointing out that society does not demand evidence of the effectiveness of arrest in crime reduction for burglaries, stranger rapes, and other nonintimate crimes (see, for example, Zorza, 1992; Frisch, 1992; Stark, 1993). Second, arrest is thought to provide a window of opportunity for a victim, since removing the offender forcibly from the household permits her to assess her options, access support and information, seek medical help, and make plans for her short-term safety and survival. Third, arrest has been hypothesized to deter offenders from repeat violence; this is probably the most familiar argument in favor of more arrests and tougher arrest policies.

The first of these hypotheses almost defies evaluation: If arrest is considered and experienced as a justly deserved punishment, then by definition it is. It is not known whether perpetrators experience arrest that way, nor is it yet known whether jurisdictions with stronger arrest policies help shape public opinion about the unacceptability of violence.6 Likewise, the second hypothesis—that arrest gives victims a needed opportunity to plan and protect themselves—is superficially plausible, but while some evidence shows that victims who call the police are more satisfied when arrests are made than when they are not (Sullivan et al., 1992a), it is not yet known if, from victims’ perspectives, arrest is a useful or critical incapacitative strategy at a moment of crisis.

The third hypothesis—that arrest deters future violence—has been the subject of numerous experimental and nonexperimental assessments. While the original and widely publicized Minneapolis experiment is still sometimes cited as authoritative evidence that arrest deters (for

6 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

example, Hart, 1996), in fact the scientific findings are much more ambiguous about the association between arrest and recidivism, and the authors of the original study have expressed concern that the findings of that study were so quickly and uncritically embraced by policymakers and administrators. The six followup studies funded by the National Institute of Justice constituted one of the most intensive and systematic investigations of a criminal justice policy question in recent decades. Experts agree that these studies offer less support for the deterrence hypothesis than the findings of the Minneapolis project seemed to promise, however.7 The relationship between arrest and future offending was associated with pretrial detention, offender characteristics such as employment, and community. In no community studied did arrestees demonstrate a dramatically different reoffense rate than did other suspects.8 (See Hirschel and Dawson, 2000, for an indepth review of the findings of these studies and their applications for policy and practice.)

Notwithstanding the limitations of the original (Minneapolis) arrest experiment, and the inconsistent results of the six replication experiments, the initial findings of that study, which seemed to offer modest support for the deterrence hypothesis, were quickly pressed into service in support of presumptive and mandatory arrest policies at both the community and State levels. However, few evaluations of mandatory arrest policies have been undertaken. It is difficult to draw conclusions now about whether mandatory arrest policies are advisable, or even whether they have the medium-run impacts that their sponsors expected, in large part because it is so difficult to compare what happened before the policy with what happened after, and in part because such policies are often adopted as part of a package of reforms whose individual effects cannot be empirically disentangled.9

The small number of evaluations that have been conducted indicate that, consistent with other criminal justice reform efforts, mandates are interpreted as advisories by practitioners. In a Midwestern State, for example, a mandatory arrest policy initially led to higher arrest rates, which then dropped after prosecutors and courts started filtering out “weak” cases (Martin, 1997). A Connecticut law produced a not-altogether-unexpected problem when police began arresting both parties on cross-complaints (Martin, 1994), clearly inconsistent with the intent of the legislation.

Arrest is not the only, or maybe even the most revealing, measure of law enforcement activity. For example, the quality of police report writing and investigation may spell success or failure for evidence-based prosecution initiatives. Police departments that train officers to thoroughly document incidents, collect evidence, take depositions, and question witnesses are likely to contribute to both probable cause findings and convictions. While such activities may be routine in serious felonies, they are also equally important in misdemeanors, where evidence is less visible. However, the impact of the quality of police investigation and report writing has seldom been researched. As practitioners recognize the importance of both documenting histories of violence and building cases from physical evidence and witness reports (not just victim allega- tions), it becomes imperative that more research be conducted on the on-scene reports that constitute the most significant documentation of a violent incident.

7 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Prosecutorial Policies and Practices: Case Retention and Innovative Prosecution Prosecutors have been subject to fewer mandates, and also less research, than police departments, probably because their discretionary decisions are less visible and less subject to litigation than are those of police. The proliferation of no-drop policies illustrates a paradox in criminal processing of domestic violence cases: In accepting responsibility for aggressively prosecuting batterers, prosecutors immediately encounter the problem of reluctant victim-witnesses, and impose a legal remedy on that problem that involves requiring victims to testify or otherwise participate, or at a minimum denies them the opportunity to request that charges be withdrawn. Critics argue that prosecutors who simply wish to spare victims the apparent responsibility for the prosecution (by taking complete charge of it themselves) have other options, although there is little evidence that they use them. A nationwide survey of urban prosecutors revealed that while two-thirds claimed some variant of a no-drop policy in domestic cases, the most common strategy for implementing this practice is to subpoena reluctant victims (Rebovich, 1996).

Evaluations of the impacts of no-drop policies are rare (see Ford and Breall, 2000, for a comprehensive discussion of research on prosecutors’ practices). The objectives of such policies might include higher case retention rates, higher conviction rates, and perhaps lower recidivism. To test the hypothesis that such a policy was effective, one would have to compare similar communities with different prosecutorial policies, or compare case processing patterns in a district attorney’s office before and after the adoption of such a policy. One of the few evaluations of this nature compared a “hard” no-drop policy (which offered victims no choice) with a “soft” policy that provided victims with an exit from prosecution. That study concluded that a soft no-drop policy was actually associated with better outcomes (regarding victim safety and recidivism) than a strict policy (Ford and Regoli, 1993).

Two other prosecutorial practices deserve mention, although research evidence on their impacts is virtually nonexistent. Diversion practices, through which prosecutors sidetrack cases away from conviction, are common in some jurisdictions for misdemeanors, particularly those that involve first offenders. Sometimes diversion involves conditions (such as compliance with an order of protection, or attendance at counseling sessions). Whether suspects who are diverted in this way behave differently from those who are convicted is an important but unresearched question. Second, evidence-based prosecution strategies (sometimes termed “victimless prosecution”), which rely on police investigations to provide independent evidence sufficient for conviction, are intended to liberate both the prosecutor and the victim from reliance on the victim’s active participation in the legal process. These kinds of strategies demand more resources and more cooperation across agencies than is customary in misdemeanor cases, but hold promise for reorienting the way judges and defense lawyers think about domestic violence cases. However, apparently few prosecutors have fully adopted this approach to domestic violence cases (Rebovich, 1996).

Court Decisions: Orders of Protection, Dispositions, and Sanctions Court decisions have been the least researched aspect of criminal processing decisions. This may be because initial problem definitions focused attention on the police, but also because historically so few cases make it as far as the courthouse. But it is primarily the courts that are

8 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

authorized to impose punitive as well as rehabilitative sanctions, and judges are largely responsible for constructing the short-term and long-term legal protections individual victims need, such as arrest warrants when offenders have fled from police, orders of protection, and conditions of offenders’ community supervision.

Postarrest sanctions and supervision are rarely imposed in misdemeanor cases, which constitute the overwhelming majority of partner violence charges. As with arrest, prosecution, and conviction, the interesting (and still unanswered) question is not, “How many suspects are ultimately punished through jail or probation?” but rather, “How much do sanctioning rates vary across communities, why, and with what effects?” Likewise, courts issue protection orders at widely differing levels; a recent study of five courts revealed that some judges issued such orders in less than 10 percent of all cases, while others issued them almost 80 percent of the time (Worden, 2000a). Court-ordered counseling for batterers remains the exception rather than the norm; even in Minneapolis, a resourceful city that has experimented with comprehensive domestic violence reforms, less than 30 percent of offenders arrested for domestic violence incidents are directed to counseling (Syers and Edleson, 1992).

A question of great interest to many practitioners—whether harsher sanctions (in the form of jail sentences) are associated with lower recidivism rates—has seldom been studied. However, one of the few research studies on this topic reached a tentative positive conclusion (Thistlethwaite, Wooldredge, and Gibbs, 1998). That study compared 1-year rearrest rates of misdemeanor offenders who received various sets of sanctions (fines, probation, and combinations of fines, probation, and jail sentences), and found that offenders who received probation and/or jail sentences were less likely to recidivate than those who received only fines or dismissal. How- ever, raising the stakes for defendants by institutionalizing tougher sanctions may have undesired (but predictable) consequences. The authors of a rare study of the impact of mandatory jail terms for some domestic violence misdemeanors concluded that the mandate to impose the terms resulted in more jail sentences, but also more plea bargaining to avoid them, which resulted in a significant drop in convictions, greater reliance on victim testimony, and longer case-processing times (Carlson and Nidey 1995). (See Worden, 2000b, for a more extensive discussion of these topics, and Saunders and Hamill, 2003, for a discussion of the impact of court-ordered treatment programs for batterers.)

One of the most important court decisions in domestic violence cases involves protection orders. While some research indicates that victims believe protection orders are helpful (Finn, 1991), other research suggests that victims are not convinced that they are enforced (Harrell, Smith, and Newmark, 1993). Recent research suggests that the efficacy of protection orders may be largely dependent on how well they are crafted and enforced, although some evidence shows that they may deter offenders under some circumstances (Keilitz, 1994; Keilitz, Hannaford, and Efkeman, 1998). It is important for practitioners to remember that in many jurisdictions, petitioners are required to seek protection orders in civil, not criminal courts, where enforcement and sanctions may be limited (Worden, 2000b).

9 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Summary These findings lead to a general and important point: The efficacy of many innovations may be contingent on the consistency of the messages that are exchanged among victims, offenders, and practitioners. Protection orders are valuable insofar as they are enforceable. Evidence-based prosecution is viable to the extent that police investigations reliably produce solid evidence of crimes. Arrest warrants may deter, but only if police officers seek them and criminal court judges issue them. The effects of criminal justice interventions may be cumulative.

This realization has led communities (and State agencies) to invest in strategies that highlight the importance of links across criminal justice agencies. Information tracking systems, specialized domestic violence courts, and cross-training on domestic violence issues across agencies (inside and outside criminal justice) have been adopted in some communities and are being contem- plated in others. These innovations sound promising in that they incorporate a more complex picture of the problem of domestic violence and impose somewhat higher expectations on community agents to craft (and cooperate on) responses.

All the same, an honest review of the research on criminal justice interventions in domestic violence reveals that many innovations produce qualified and incomplete answers to the urgent question, “What works?” At this point, research neither justifies elimination of currently popular innovations nor supports recommendations for wholesale adoption of most strategies. Like the specific practices and programs briefly reviewed above, coordinated community responses have been enthusiastically endorsed by funding agencies, victim advocates, and criminal justice practitioners as a vehicle for incorporating a working understanding of violence and victimization into an operational criminal justice response. The following sections review what is known about the structure and functions of coordinated responses, the particular problems practitioners and researchers face in figuring out how and why coordinated community initiatives might work best, and what has been learned so far about their effect on outcomes.

Coordinated Community Responses to Domestic Violence: Evaluations of Process and Outcomes

Many community-level activities involving diverse sets of actors and agencies have been included under the heading “coordinated community responses.” Hart (1996) distinguishes these projects by structure and participation (for example, limited partnerships between domestic violence programs and specific criminal justice agencies, such as those sponsored by the Office of Community Oriented Policing Services); comprehensive interventions run by nonprofit independent agencies (for example, the Duluth Domestic Abuse Intervention Program [Shepard and Pence, 1999]); and externally (State or Federal) sponsored technical assistance and training programs (Wilder et al., 1995). These activities often are initiated by a local task force or coalition whose mission is to identify community problems in responding to domestic violence, formulate solutions, and develop support for implementing those solutions (Hart, 1996).

10 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Community coordination efforts vary in other ways as well. Although victim advocacy programs often initiate task forces, criminal justice agents and/or advocates may provide leadership. Participation may be limited to particular agencies, may be broadly inclusive of agencies with victim and offender contact, or may reach out to community organizations that might engage in prevention and education as well as intervention (such as schools, churches, and health care providers). While case studies of communities illustrate the diversity of strategies and activities, however, it is not known how common different sorts of arrangements are.10

Moreover, published studies of the implementation (process) and impacts (outcomes) of interagency initiatives offer limited insight into the circumstances under which they are undertaken and successfully sustained, or whether they achieve measurable changes in stated objectives. Although these initiatives are proliferating, little is known about how, and how well, they work. In fact, a recent comprehensive assessment of research on family violence concluded that coordinated community responses to domestic violence had not yet been the subject of outcome evaluations that met conventional scientific standards (Chalk and King, 1998). This is true for several reasons:

Initiatives are often well underway before process or outcome evaluations are undertaken, so no opportunity exists for researchers to gather data on system performance before the new project.

Observations of single sites cannot exclude idiosyncratic local factors as explanations for particular outcomes, yet comparisons of several sites working in similar conditions are difficult and expensive.

Cross-agency efforts are more problematic and challenging and require more time and energy to implement than in-house policy directives; moreover, the goals of such efforts are often commensurately more ambitious and long term, but evaluations must be completed in too brief a time to draw confident conclusions about the routinization of change or its impacts (O’Connor, 1995; Brown, 1995).

Ethical concerns sometimes conflict with research strategies. For example, police, prosecutors, judges, and service providers may be uncomfortable implementing experimental designs that randomly assign treatment, punishment, or other interventions to offenders (or services to victims), especially if they have opinions about the fairness or effectiveness of particular interventions (Sherman et al., 1992; Feder, 1998). Researchers can also confront ethical issues when they try to gather information from victims and offenders, including issues of privacy, safety, and intrusiveness.

Given limited resources, researchers are naturally inclined to study programs and projects that appear particularly promising (and perhaps there is a natural tendency among such projects’ leadership to be receptive to evaluation), and as a result published research probably overrepresents the incidence of successfully implemented and operating coordination efforts. Evaluation research has produced far fewer assessments of initiatives that have

11 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

failed, fizzled, or simply become dormant, even though those communities’ experiences might teach a great deal about the factors that are important to creating and sustaining effective projects.

With these caveats in mind, the following sections summarize what is known so far from both process and outcome evaluations of various types of coordinated responses to domestic violence.

Evaluating the Adoption and Implementation of Coordinated Responses Several well-known studies of implementation of community coordination projects have been published, and they tell different tales about the conditions that promote successful implementation. Duluth (Pence, 1983; Shepard and Pence, 1999) and Minneapolis, Minnesota (Brygger and Edleson, 1987; Gamache, Edleson, and Schock, 1988), and Santa Barbara, California (Berk et al., 1982), have been the sites of coordinated intervention efforts organized by nonprofit agencies and aimed at increasing offender accountability and improving the consistency of criminal justice responses. The goals of the Duluth project were to coordinate criminal justice and human services (including police, local corrections, prosecutors, victim services, and independent counseling agencies) toward shared objectives: redefining battering as a crime, shifting the focus of intervention to offenders, and ultimately reducing the number of victims experiencing repeat violence. The emphasis was initially on increasing the use of criminal justice outcomes at all stages of the process and reorienting existing counseling programs toward batterer accountability. An early process evaluation reported success in many of these objectives—success which was attributed to program staff’s strong commitment to coordinating, rather than reinventing, the local response from criminal justice and victim services practitioners (Pence, 1983).

Similarly structured efforts in Minneapolis had similar objectives (Brygger and Edleson, 1986), replicating a commitment to ending violence in relationships through intensive intervention with batterers. A multisite evaluation of compliance with project goals, which measured increased arrest rates, prosecution rates, and use of mandated counseling, revealed positive results in a 2- year study, although the authors did not examine the impacts of criminal justice actions on recidivism (Gamache, Edleson, and Schock, 1988).

However, a cautionary note was sounded in an early study published by Berk et al. (1982). An initial evaluation of a family violence program in Santa Barbara produced cause for optimism: Within a year of their adoption, new criminal justice policies to improve police reporting, arrests, sentencing, and mandatory counseling appeared to have produced desired results (Berk, Berk, and Rauma, 1980). However, a followup evaluation after a second year painted a much different picture, as the increases in reports written, sentences imposed, and use of counseling had sunk to preinitiative levels (Berk et al., 1982). In what could best be described as a postmortem, the authors reported that the project had lost the support of rank-and-file police officers, who resisted imposed training and ultimately had it abolished, and had suffered with the resignation of the police chief and the withdrawal of Law Enforcement Assistance Administration funds from the prosecutor’s office.

12 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Lincoln, Nebraska, was the site of a somewhat different model of intervention during the late 1980s. A criminal justice task force undertook an effort to coordinate official actions in an attempt to maximize case retention and reduce recidivism. A study of police reporting and arrest policies before and after the official start of the project did not reveal significant changes (Steinman, 1991).

Taken together, these studies should recommend caution to practitioners and policymakers who rely primarily on the availability of resources, particularly external funding, to jumpstart and sustain new programs. Research is not adequate to allow generalizations about the circumstances under which coordination efforts are likely or unlikely to be adequately implemented, even by their own stated standards. Even when such standards are quite explicit and seemingly proximate to practitioner behavior and supervision (such as increasing report writing or arrest rates in domestic cases), change is difficult to achieve.

Two standards tentatively appear to be common to successful community change, however. Initiatives that build on existing community leadership (as opposed to outsiders) and that delegate responsibility to local practitioners may have a greater chance of surviving. Also, although victims are most often the subject of discussion about reform, projects that are aimed at getting and keeping perpetrators in the criminal justice process have proven feasible under some conditions; this is an important observation if one defines offender accountability in terms that include but are not limited to victim safety.

Evaluating the Impacts of Coordinated Responses Researchers have had somewhat more success, or at least more experience, in assessing the impacts of coordinated interventions, although again, the research offers limited purchase for generalization to other sites. Almost all systematic research on outcomes has focused on offender recidivism. Many studies have examined the differences in reported recidivism rates for offenders who are subjected to various experiences (arrest, counseling, etc.). The distinctive element of coordinated strategies, however, is the potential to subject offenders to multiple and cumulative legal interventions. Most empirical evaluations of these interventions are limited insofar as they involve small numbers of offenders (so it is difficult to generalize to larger populations) or do not statistically control for factors that might account for behavioral changes (for example, judges might be more likely to sentence first offenders to counseling than offenders with prior records, but first offenders might be less likely to reoffend for reasons unrelated to counseling).

The Minneapolis Intervention Project assessed recidivism after a year among three groups: men who were not arrested following a domestic violence call, men who were arrested but not sanctioned, and men who were arrested and mandated to treatment. Recidivism rates were high, averaging 40 percent for all three groups (based on victim interviews and police records). They were slightly lower for men mandated to counseling, although differences in recidivism among these groups were not found to be statistically significant by the authors (Syers and Edleson, 1992). Although that study was limited by small sample sizes, the authors tentatively concluded that arrest followed by long-term monitoring in treatment may be associated with lower

13 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

recidivism. A similar but more complex study of interventions in Duluth’s program attempted to predict recidivism among a population of charged offenders using measures of criminal justice and sentencing experiences (chemical dependency evaluations, involvement in civil versus criminal court, participation in counseling and education sessions) as well as personal back- ground factors. The author concluded that personal factors (specifically chemical dependency, history of child abuse victimization, and previous convictions for nonviolent crimes) were better predictors of future violent behavior than the specific interventions imposed on offenders (Shepard, 1992).

Several studies have focused on the cumulative impacts of criminal justice interventions under community coordination projects, implicitly or explicitly hypothesizing that the effects of apprehension, prosecution, and sentencing, including monitoring, add up—that offenders exposed to more interventions or more diverse legal reactions are less likely to recidivate. However, evidence supporting this general hypothesis is mixed. Steinman (1991) found few differences in recidivism after police and prosecutors coordinated their activities in an attempt to increase case retention. Little evidence showed that any combination of sanctions (probation, counseling, or fines) was associated with recidivism in Lincoln, Nebraska (Steinman, 1988). A similar program in DuPage County, Illinois, also aimed at increasing case retention and increasing the use of the criminal process, found that arrested offenders were somewhat less likely than those who were not arrested on the scene to recidivate, although no differences were found in recidivism rates among those in the arrested group who were and were not ultimately convicted of an offense (Tolman and Weisz, 1995).

These studies were conducted in communities that are comparatively middle class, with stable employment rates and relatively homogeneous populations. A more recent study in Baltimore (Murphy, Musser, and Maton, 1998), which included a more typical urban population of offenders, explicitly examined the association between recidivism and specific criminal justice actions (arrest, prosecution, orders of protection, counseling, and sentencing) and between recidivism and combinations of these responses, controlling for individual characteristics such as prior record, seriousness of the offense, race, age, and legal representation. The authors found that although no single legal response was associated with recidivism, the accumulation of responses was significantly associated with reduced recidivism over a 1-year period. This study is one of the most explicit tests of a core tenet of most coordinated criminal justice responses: the belief that a criminal justice system that predictably and routinely entangles offenders in multiple ways improves the odds that any given offender will encounter a response that may alter his behavior.

One important cautionary note is in order. Most of these studies focus primarily on recidivism; not surprisingly to most researchers and practitioners, few reveal significant changes in or differences among recidivism rates over time or across groups exposed to different criminal justice environments. This may be due in part to the difficulty of accurately measuring recidivism, especially when the only means available is to count on victims’ repeat calls to police. It may also simply be that violence in relationships is an uncommonly difficult behavior to change, so “one shot” responses are unlikely to make much difference. Alternative conceptualizations of

14 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

desirable outcomes—increases in victim reporting, for example—are likely to be the target of future research (see, for example, Davis and Taylor, 1997).

Summary and Conclusions

This report intends to summarize usable, research-based information on questions and problems that confront people who are working to better coordinate the response of community agencies, especially criminal justice agencies, to domestic violence. Coming from different professional backgrounds and responsible for different constituencies, participants in local task forces confront a seeming landslide of information, recommendations, and opportunities for change. The pace of innovation and rising expectations may have outrun the pace of careful, knowledge- based planning and certainly has outstripped most practitioners’ ability to stay abreast of research findings. As a result, task force members might benefit from sharing what researchers have learned about conventional wisdom regarding domestic violence and criminal justice, the efficacy of specific reforms that are often on local agendas, and what is known about the conditions that promote and sustain coordination efforts and the types of community changes that are associated with desired outcomes.

Research shows, first, that generalizations about police, courts, offenders, and victims are often unreliable because they oversimplify and overgeneralize complex decisions and behaviors. Second, behaviors often characterized as constants are in fact variables; arrest rates, conviction rates, recidivism risk, and victim preferences and behavior vary at both the individual and the community levels. What is interesting, therefore, is learning what accounts for that variation; if practitioners want to change what happens in their own communities, they stand to learn a great deal from the experiences of other jurisdictions. Research is only beginning to examine these kinds of variables at the community level.

Third, research reveals that many widely adopted reforms have limited, and sometimes condi- tional, effects on outcomes. Arrest has been the most frequently studied policy change, but not enough is known about the effects of arrest—on offender behavior as well as on victim reporting behavior and victim safety—to confidently endorse mandatory arrest laws (Sherman et al., 1992). Likewise, what is known so far about prosecutor and court reforms is inconclusive. Ideas that sound good in principle may have unintended consequences; existing policies (when practiced by informed and well-trained professionals) may be as beneficial for victims and offenders as policies that reduce or limit discretion. Research that focuses exclusively on official criminal justice outcomes (such as recidivism as measured by rearrest, or compliance with probation conditions) sometimes overlooks important outcomes of concern to victims and victim advocates.

Community coordination efforts emerged from local agents’ realization that piecemeal changes in policy or practice (which may stem from simplistic assumptions about the causes of the problem) seldom provide hoped-for results. This led to the realization that the problem may rest in the lack of consistency in responses from many sectors of the community, and eventually, thoughtful observers recognized that a more sophisticated and informed understanding of the

15 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

problem must be incorporated into the solution, which itself becomes more complex and multifaceted. In the case of responses to domestic violence, initial problem definitions that target particular points in the criminal process yielded more complex definitions that call for coordinated responses from multiple agents.

Researchers and practitioners face two important empirical questions: “What factors are necessary or sufficient to support a coordinated response?” and “Do coordination projects produce better results?” Research suggests that it might be easier to answer the first question in the obverse: “What factors detract from such efforts?” Resistance from local police (at leadership and line levels) and attempts to induce dramatic changes in practice seem unlikely to succeed. On the other hand, coordination strategies that are not dependent on any particular agency’s leadership or ownership, but that build on existing resources and relationships, may have a better chance of success. Case studies suggest that leadership and the presence of a committed local core of participants may be vital.

The answer to the second question depends on how one measures results, and most measures have been conventional criminal justice outcomes. But a review of research literature offers little evidence that any particular strategy is associated with dramatic changes in conventionally measured outcomes. More is still being learned about outcomes that have typically gone unobserved: victim reporting decisions, victims’ levels of trust in and reliance on community and criminal justice agents, and offenders’ behavior that is legal but nonetheless constitutes abuse and may be a good predictor of future physical violence. There is much left to learn.

This summary observation sounds pessimistic at first: Existing research does not show that any particular strategy makes a significant difference. However, there is a more optimistic interpreta- tion: These evaluations have illuminated the unintended consequences of changes and illustrated the limits of the criminal justice process in bringing about real change at the individual and community levels. In doing so, they have highlighted the importance of investing more carefully in selecting outcome measures, defining outcomes at both levels and comparing experiences across communities to learn what political, cultural, and agency variables might be associated with more and less effective results.

Notes

1. Coordinated community interventions are not new to criminal justice or unique to the problem of violence against women. Since the 1960s, communities have been the locus of many organized responses to social problems, including teen pregnancy, youth unemployment, and crime prevention (Kubisch et al., 1995). Communities appear to be the most promising target of change efforts for many reasons. Community leaders are knowledgeable about and can capitalize on existing governmental and nongovernmental institutions; the social organization of communities is often identified as a source of the problem; and the limits of intervention are bounded by geographical lines that have social and legal meaning (O’Connor, 1995). This is especially true of problems involving crime and criminal justice: Local criminal justice systems

16 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

are unique, characterized by idiosyncratic practices and customs, and their jurisdictions are clearly bounded by precinct, city, and county lines.

2. Formally, laws that prohibit various acts of violence against women vary tremendously. While some States’ statutes include express prohibitions on types of assault or other crimes for family members, other States simply prohibit assault and have stopped using common-law exceptions for wives. Often, although not always, these laws specify the nature of legal relationships between offender and victim that are covered by the statute, so they vary in their applicability to unmarried, cohabiting, dating, and same-sex partners. In practice, community agencies, including criminal justice agencies, may explicitly or implicitly adopt definitions of violence for the purposes of applying local practices or protocols, or for program eligibility, that may be more inclusive or exclusive than criminal law provisions.

3. A recent national survey-based study concluded that “[n]early 25 percent of surveyed women and 7.6 percent of surveyed men had been raped and/or physically assaulted by a current or former spouse, cohabiting partner, or date at some time in their lifetime” (Tjaden and Thoennes, 2000, p. iii) and also that almost 5 percent of women had been stalked by a partner or former partner (Tjaden and Thoennes, 2000); see Bachman, 2000, for a detailed discussion of prevalence and reporting rates.

4. The traditional standard for misdemeanor arrest requires that the officer witnessed the offense, procure a warrant to justify arrest, or have a citizen sign a complaint; the “warrantless arrest” provision of many States’ laws was intended to expedite arrests and remove the responsibility from victims for formally filing charges against their abusers (Zorza, 1992).

5. Researchers have also examined the factors that are associated with the decision to make an arrest in a domestic incident. Generally, the research indicates that legal factors (seriousness of offense) and the presence of physical evidence are associated with the arrest decision, while characteristics of the victim and offender are not associated with outcomes consistent with most other studies of arrest (Berk and Loseke, 1980; Worden and Pollitz, 1984). However, since widespread statutory and departmental reforms were adopted in the late 1980s, little research has been published based on independent observation of police behavior rather than on the official data generated by police reports.

6. This is not to say that these questions could not be researched. The Spouse Assault Replication Program studies included offender interviews, and while those interviews were incomplete and conducted under less than ideal circumstances, they might offer some insight into how chastened offenders felt by the experience of arrest—although because offenders who were not arrested were not interviewed, the specific effect of arrest would be impossible to measure (see Paternoster et al., 1997). Enough jurisdictions have experimented with stronger arrest policies now to permit comparative studies of public opinion under these differing legal cultures.

7. Most of the studies in the experimental series were in fact analyzing the repeat arrest (or callback) rate of offenders who not only had prior arrest records, but also had histories of partner violence, so it is risky to draw conclusions about the effects of a particular isolated experience in

17 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

what, for most offenders, was a rather long series of violent incidents and encounters with police. Thoughtful observers also have pointed out that to the extent studies rely on police data for evidence of recidivism, they may really be explaining victim behavior: Since most victims do not report most violent behavior, it may be as logical to attribute callbacks to victims’ decisions to report as to the actual incidence of violence. Finally, an often overlooked feature of these studies is that arrest was typically one of several “responses,” which included options such as on-scene mediation, separation of the parties for the night, and even an implied threat to arrest one or both parties if the police were called back; it is worthwhile to notice that researchers did not find pronounced effects of arrest on recidivism, compared with most of these options.

8. A unique and somewhat unanticipated finding of the Omaha arrest study (Dunford, Huizinga, and Elliott, 1990) was that the issuance of an arrest warrant (even one that was unlikely to be executed) was associated with lower recidivism rates. This is one of the few findings in the series of studies that actually supports a deterrence tenet: By issuing warrants, officers increased the probability of sanctions on reoffending.

9. For example, the express goal of New York’s 1994 mandatory arrest policy was to reduce recidivism in family violence situations. However, the only practical way to measure recidivism at the State level is through police contacts or rearrests; if indeed police are arresting invariantly now, there is no way to assess whether arrest rates are a better reflection of incidence rates, of police compliance with the statute, or neither.

10. For an indepth review of the efforts of six model communities’ coordination responses to domestic violence, see Clark et al., 1996.

References

Albonetti, C.A. “Prosecutorial Discretion: The Effects of Uncertainty.” Law and Society Review, 21: 291–313 (1987).

Bachman, R. “Violence Against Women: Synthesis of Research for Criminal Justice Policy- makers.” Final Report, grant number 98–WT–VX–K011, submitted to the U.S. Department of Justice, National Institute of Justice (2000). NCJ 199579.

Bard, M., and J. Zacker. “The Prevention of Family Violence: Dilemmas of Community Interac- tion.” Journal of Marriage and the Family, 33: 677–682 (1971).

Berk, R.A., S.F. Berk, and D. Rauma. “Bringing the Cops Back In: A Study of Efforts to Make the Criminal Justice System More Responsive to Incidents of Family Violence.” Social Science Research, 9: 193–213 (1980).

Berk, S.F., and D. Loseke. “‘Handling’ Family Violence: Situational Determinants of Police Arrest in Domestic Disturbances.” Law and Society Review, 15: 317–346 (1980).

18 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Berk, R.A., D. Rauma, D. Loseke, and S.F. Berk. “Throwing the Cops Back Out: The Decline of a Local Program to Make the Criminal Justice System More Responsive.” Social Science Research, 11: 245–279 (1982).

Bonsignore, J.J., E. Katsh, P. d’Errico, R. Pipkin, S. Arons, and J. Rifkin (eds.). Before the Law: An Introduction to the Legal Process. 4th ed. Boston: Houghton Mifflin (1989).

Brown, P. “The Role of the Evaluator in CCIs.” In New Approaches to Evaluating Community Initiatives, J.P. Connell, A. Kubisch, L.B. Schorr, and C.H. Weiss, eds. Washington, DC: Aspen Institute (1995).

Brygger, M.P., and J.L. Edleson. “The Domestic Abuse Project: A Multisystems Intervention in Woman Battering.” Journal of Interpersonal Violence, 2(3): 324–336 (1987).

Buzawa, E., and C. Buzawa. Domestic Violence: The Criminal Justice Response. Thousand Oaks, CA: Sage Publications (1996).

Buzawa, E., G. Hotaling, and A. Klein. “The Response to Domestic Violence in a Model Court: Some Initial Findings and Implications.” Behavioral Science and the Law, 16: 185–206 (1998).

Carlson, C., and F.J. Nidey. “Mandatory Penalties, Victim Cooperation, and the Judicial Processing of Domestic Abuse Assault Cases.” Crime and Delinquency, 41: 132–139 (1995).

Chalk, R., and P. King. Violence in Families: Assessing Prevention and Treatment Programs. Washington, DC: National Academy Press (1998).

Clark, S.J., M.R. Burt, M.M. Schulte, and K. Maguire. Coordinated Community Responses to Domestic Violence in Six Communities: Beyond the Justice System. Washington, DC: Urban Institute for the Office of the Assistant Secretary of Planning and Evaluation (1996). http://aspe.os.dhhs.gov/hsp/cyp/domvilnz.htm

Davis, R.C., F. Kunreuther, and E. Connick. “Expanding the Victim’s Role in the Criminal Court Dispositional Process: The Results of an Experiment.” Journal of Criminal Law and Criminol- ogy, 75: 491–505 (1984).

Davis, R.C., B.E. Smith, and L.B. Nickles. “The Deterrent Effect of Prosecuting Domestic Violence Misdemeanors.” Crime and Delinquency, 44: 434–442 (1998).

Davis, R.C., and B. Taylor. “A Proactive Response to Family Violence: The Results of a Randomized Experiment.” Criminology, 35(1): 307–333 (1997).

Dunford, F.W. “System Initiated Warrants for Suspects of Misdemeanor Domestic Assault: A Pilot Study.” Justice Quarterly, 7: 631–653 (1990).

19 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Dunford, F.W., D. Huizinga, and D.S. Elliott. “The Role of Arrest in Domestic Assault: The Omaha Police Experiment.” Criminology, 28(2): 183–206 (1990).

Dutton, D.G. The Domestic Assault of Women: Psychological and Criminal Justice Perspectives. Boston: Allyn and Bacon (1988).

Erez, E. “Intimacy, Violence and the Police.” Human Relations, 39: 265–281 (1986).

Erez, E., and J. Belknap. “In Their Own Words: Battered Women’s Assessment of the Criminal Processing System’s Responses.” Violence and Victims, 13: 251–268 (1998).

Fagan, J. “Cessation of Family Violence: Deterrence and Dissuasion.” In Family Violence, L. Ohlin and M. Tonry, eds. Chicago: University of Chicago Press (1989).

Feder, L. “Police Handling of Domestic Violence and Police Response in a Pro-Arrest Jurisdic- tion.” Crime and Delinquency, 44: 335–349 (1998).

Ferraro, K.J. “Policing Woman Battering.” Social Problems, 36: 61–74 (1989).

Finn, P. “Civil Protection Orders: A Flawed Opportunity for Intervention.” In Woman Battering: Policy Responses, M. Steinman, ed. Cincinnati: Anderson Publishing Company (1991).

Ford, D.A. “Prosecution as a Victim Power Source: A Note on Empowering Women in Their Violent Conjugal Relationships.” Law and Society Review, 25: 313–334 (1991).

Ford, D.A., and S. Breall. “Violence Against Women: Synthesis of Research for Prosecutors.” Final Report, grant number 98–WT–VX–K011, submitted to the U.S. Department of Justice, National Institute of Justice (2000). NCJ 199660.

Ford, D.A., and M.J. Regoli. “The Indianapolis Domestic Violence Prosecution Experiment.” Final Report, grant number 86–IJ–CX–0012, submitted to the U.S. Department of Justice, National Institute of Justice (1993). Indianapolis: Indiana University. NCJ 157870.

Ford, D.A., and M.J. Regoli. “The Effectiveness of Defense Counsel in Cases of Wife Battery.” Paper presented at the Law and Society Association Meeting in Berkeley, California (1990).

Frisch, L.A. “Research that Succeeds, Policies that Fail.” Journal of Criminal Law and Criminol- ogy, 83: 209–216 (1992).

Frohmann, L. “Convictability and Discordant Locales: Reproducing Race, Class, and Gender Ideologies in Prosecutorial Decision Making.” Law and Society Review, 31: 531–555 (1997).

Fyfe, J.J., D.A. Klinger, and J.M. Flavin. “Differential Police Treatment of Male-on-Female Spousal Violence.” Criminology, 35(3): 455–473 (1997).

20 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Gamache, D.J., J.L. Edleson, and M.D. Schock. “Coordinated Police, Judicial, and Social Service Response to Woman Battering: A Multiple Baseline Evaluation Across Three Communi- ties.” In Coping with Family Violence: Research and Policy Perspectives, G.T. Hotaling, D. Finkelhor, J.T. Kirkpatrick, and M.A. Straus, eds. Newbury Park, CA: Sage Publications (1988).

Garner, J., J. Fagan, and C. Maxwell. “Published Findings From the Spouse Assault Replication Program: A Critical Review.” Journal of Quantitative Criminology, 11: 3–28 (1995).

Harrell, A., B. Smith, and L. Newmark. Court Processing and the Effects of Restraining Orders for Domestic Violence Victims. Washington, DC: Urban Institute (1993).

Hart, B.J. “Battered Women and the Criminal Justice System.” In Do Arrests and Restraining Orders Work?, E. Buzawa and C. Buzawa, eds. Thousand Oaks, CA: Sage Publications (1996).

Hirschel, J.D., and D.J. Dawson. “Violence Against Women: Synthesis of Research for Law Enforcement Officials.” Final Report, grant number 98–WT–VX–K011, submitted to the U.S. Department of Justice, National Institute of Justice (2000). NCJ 198372.

Hirschel, J.D., and I.W. Hutchison III. “Female Spouse Abuse and the Police Response: The Charlotte, North Carolina Experiment.” Journal of Criminal Law and Criminology, 83: 73–119 (1992).

Keilitz, S. “Civil Protection Orders: A Viable Justice System Tool for Deterring Domestic Violence.” Violence and Victims, 9: 79–84 (1994).

Keilitz, S., P. Hannaford, and H.S. Efkeman. “The Effectiveness of Civil Protection Orders.” In Legal Interventions in Family Violence, Research Findings and Policy Implications, American Bar Association. Chicago: American Bar Association (1998).

Kubisch, A.C., C.H. Weiss, L.B. Schorr, and J.P. Connell. “Introduction: New Approaches to Evaluating Community Initiatives.” In New Approaches to Evaluating Community Initiatives, J.P. Connell, A.C. Kubisch, L.B. Schorr, and C.H. Weiss, eds. Washington, DC: Aspen Institute (1995).

Martin, M.E. “Double Your Trouble: Dual Arrest in Family Violence.” Journal of Family Violence, 12: 139–157 (1997).

Martin, M.E. “Mandatory Arrest for Domestic Violence: The Courts’ Response.” Criminal Justice Review, 19(2): 212–227 (1994).

Merry, S.E. Getting Justice and Getting Even: Legal Consciousness Among Working-Class Americans. Chicago: University of Chicago Press (1990).

21 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Murphy, C.M., P.H. Musser, and K.I. Maton. “Coordinated Community Intervention for Domestic Abusers: Intervention, System Involvement, and Criminal Recidivism.” Journal of Family Violence, 13(3): 263–284 (1998).

Myers, M.A., and J. Hagan. “Private and Public Trouble: Prosecution and the Allocation of Court Resources.” Social Problems, 26: 439–451 (1979).

O’Connor, A. “Evaluating CCIs: A View From History.” In New Approaches to Evaluating Community Initiatives, J.P. Connell, A. Kubisch, L.B. Schorr, and C.H. Weiss, eds. Washington, DC: Aspen Institute (1995).

Parnas, R.I. “The Police Response to the Domestic Disturbance.” Wisconsin Law Review, 31: 914–960 (1967).

Paternoster, R., R. Bachman, R. Brame, and L.W. Sherman. “Do Fair Procedures Matter? The Effect of Procedural Justice on Spouse Assault.” Law and Society Review, 3: 163–204 (1997).

Pence, E. “The Duluth Domestic Abuse Intervention Project.” Hamline Law Review, 6: 247–275 (1983).

Pleck, E. Domestic Tyranny: The Making of American Social Policy Against Family Violence From Colonial Times to the Present. New York: Oxford Press (1987).

Rauma, D. “Going for the Gold: Prosecutorial Decision-Making in Cases of Wife Assault.” Social Science Research, 13: 321–351 (1984).

Rebovich, D.J. “Prosecution Response to Domestic Violence: Results of a Survey of Large Jurisdictions.” In Do Arrests and Restraining Orders Work?, E. Buzawa and C. Buzawa, eds. Thousand Oaks, CA: Sage Publications (1996).

Reynolds, D.E. “The Use of Pre-trial Diversion Programs in Spouse Abuse Cases: A New Solution to an Old Problem.” Ohio State Journal in Dispute Resolution, 2: 415–436 (1988).

Ryan, R.M. “The Sex Right: A Legal History of the Marital Rape Exemption.” Law and Social Inquiry, 21: 941–1001 (1996).

Saunders, D.G. “A Typology of Men Who Batter: Three Types Derived from Cluster Analysis.” American Journal of Orthopsychiatry, 62: 264–275 (1992).

Saunders, D.G., and R. Hamill. “Violence Against Women: Synthesis of Research on Offender Interventions.” Final Report, grant number 98–WT–VX–K011, submitted to the U.S. Department of Justice, National Institute of Justice (2003). NCJ 201222.

22 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Schechter, S. “Toward an Analysis of Violence Against Women in the Family.” In Women and Male Violence: The Visions and Struggles of the Battered Women’s Movement. Boston: South End Press (1982).

Schmidt, J., and E.H. Steury. “Prosecutorial Discretion in Filing Charges in Domestic Violence Cases.” Criminology, 27: 487–510 (1989).

Shepard, M. “Predicting Batterer Recidivism Five Years After Community Intervention.” Journal of Family Violence, 7: 167–178 (1992).

Shepard, M., and E. Pence. Coordinating Community Responses to Domestic Violence: Lessons from Duluth and Beyond. Thousand Oaks, CA: Sage Publications (1999).

Sherman, L.W., J.D. Schmidt, D.P. Rogan, P.R. Gartin, E.G. Cohn, D.J. Collins, and A.R. Bacich. “From Initial Deterrence to Long-Term Escalation: Short Custody Arrest for Poverty Ghetto Domestic Violence.” Criminology, 29: 821–850 (1991).

Sherman, L.W., D.A. Smith, J.D. Schmidt, and D.P. Rogan. “Crime, Punishment, and Stake in Conformity: Legal and Informal Control of Domestic Violence.” American Sociological Review, 57: 680–690 (1992).

Stark, E. “Mandatory Arrest of Batterers.” American Behavioral Scientist, 36(5): 651–680 (1993).

Steinman, M. “Evaluating a System-Wide Response to Domestic Abuse: Some Initial Findings.” Journal of Contemporary Criminal Justice, 4: 172–188 (1988).

Steinman, M. (ed.). Woman Battering: Policy Responses. Cincinnati: Anderson Publishing Company (1991).

Sullivan, C., R. Campbell, H. Angelique, K. Eby, and W. Davidson. “An Advocacy Intervention Program for Women With Abusive Partners.” American Journal of Community Psychology, 22(1): 101–122 (1994).

Sullivan, C.M., J. Basta, C. Tan, and W.S. Davidson II. “After the Crisis: A Needs Assessment of Women Leaving a Domestic Violence Shelter.” Violence and Victims, 7: 267–275 (1992a).

Sullivan, C.M., C. Tan, J. Basta, M.H. Rumptz, and W.S. Davidson II. “An Advocacy Interven- tion Program for Women With Abusive Partners: Initial Evaluation.” American Journal of Community Psychology, 20: 309–332 (1992b).

Syers, M., and J.L. Edleson. “The Combined Effects of Coordinated Criminal Justice Interven- tion in Woman Abuse.” Journal of Interpersonal Violence, 7(4): 490–502 (1992).

23 Violence Against Women: Synthesis of Research for Task Forces and Coalition Members

Thistlethwaite, A., J. Wooldredge, and D. Gibbs. “Severity of Dispositions and Domestic Violence Recidivism.” Crime and Delinquency, 44: 388–398 (1998).

Tjaden, P., and N. Thoennes. Extent, Nature, and Consequences of Intimate Partner Violence: Findings From the National Violence Against Women Survey. Research Report. Washington, DC: U.S. Department of Justice, National Institute of Justice, and U.S. Department of Health and Human Services, Centers for Disease Control and Prevention (2000). NCJ 181867.

Tolman, R.M., and A. Weisz. “Coordinated Community Intervention for Domestic Violence: The Effects of Arrest and Prosecution on Recidivism of Woman Abuse Perpetrators.” Crime and Delinquency, 41: 481–495 (1995).

Wilder, S.C., P. Pahl, T. Young, J. Shenk, J. Mork-O’Brien, K. Kerston, and R. Ryan. The Washington State Domestic Violence Act: An Evaluation Project. (1995).

Williams, K., and R. Hawkins. “Wife Assaults, Costs of Arrest and the Deterrence Process.” Journal of Crime and Delinquency Research, 29: 292–310 (1992).

Worden, A.P. “Models of Community Coordination in Partner Violence Cases.” Draft report, grant number 95–WT–NX–0006, submitted to the U.S. Department of Justice, National Institute of Justice (2000a).

Worden, A.P. “Violence Against Women: Synthesis of Research for Judges.” Final Report, grant number 98–WT–VX–K011, submitted to the U.S. Department of Justice, National Institute of Justice (2000). NCJ 199911.

Worden, A.P., B.E. Carlson, J. Postmus, and M. van Ryn. “Practitioner Perspectives on Violence Against Women.” Paper presented at the annual meeting of the American Society of Criminol- ogy, Toronto (1999).

Worden, R., and A. Pollitz. “Police Arrests in Domestic Disturbances: A Further Look.” Law and Society Review, 18: 105–119 (1984).

Zorza, J. “The Criminal Law of Misdemeanor Domestic Violence, 1970–1990.” Journal of Criminal Law and Criminology, 83: 46–72 (1992).